The biggest DeFi security incident of 2026: KelpDAO cross-chain bridge vulnerability, Aave nearly $200 million in bad debt

In April 2026, the cryptocurrency industry underwent a severe systemic risk test for DeFi. The KelpDAO cross-chain bridge was attacked due to a LayerZero configuration vulnerability, which led to an abnormal increase in the issuance of rsETH, and then propagated to the Aave protocol, forming nearly $200 million in bad debt. Within 72 hours, the total value locked (TVL) across the entire DeFi ecosystem evaporated by more than $13 billion. This incident not only exposed the risk exposure between cross-chain bridges and lending protocols, but also triggered in-depth discussions about the security boundaries of composability.

How did the KelpDAO attack trigger nearly $200 million in bad debt for Aave?

The attack’s propagation path can be divided into three stages. In the first stage, the attacker exploited a LayerZero configuration vulnerability in the KelpDAO cross-chain bridge, bypassing the permission-checking mechanism and illegally minting a large amount of rsETH on the source chain. In the second stage, the attacker transferred the minted rsETH across chains to the Ethereum mainnet and quickly exchanged it for other assets on multiple DEXes, causing the rsETH price to temporarily depeg. In the third stage, because Aave integrated rsETH as a collateral asset, the attacker used the minted rsETH to over-borrow ETH and USDC and then directly withdrew liquidity, leaving behind bad debt that could not be liquidated. As of April 20, 2026, the bad-debt size disclosed by Aave was approximately $177 million to $200 million, with the exact figure depending on subsequent liquidation and recovery progress.

How did the rsETH cross-chain bridge vulnerability and LayerZero misconfiguration expose the issue?

The technical root of this attack lies in a flaw in the cross-chain bridge’s permission management. The cross-chain bridge solution adopted by KelpDAO is based on LayerZero’s general message passing protocol, but the configuration did not strictly verify the contract address of the message sender. The attacker forged the identity of a legitimate message sender and submitted a “mint” instruction to the target chain. LayerZero’s relayer and endpoint contracts executed the message normally because their verification mechanism only checked the message signature, without performing a secondary check on the business legitimacy of the message content. This vulnerability is a typical “mismatch between configuration and business logic” problem, which has appeared multiple times in cross-chain bridge attack incidents from 2025 to 2026. As rsETH is KelpDAO’s liquidity restaking token, its minting permission was originally limited to specific contracts, but the cross-chain bridge’s minting interface was mistakenly exposed to external callers.

Why couldn’t Aave avoid the $177 million to $200 million bad debt?

As a decentralized lending protocol, Aave’s risk-control model relies on on-chain oracle prices and liquidation mechanisms. In this incident, the duration of rsETH’s price depeg was extremely short, and the attacker had already completed the borrowing operations before the price fell. When the rsETH price began to drop, the attacker’s positions were already “underwater,” but Aave’s liquidation bots failed to trigger in time for two reasons. First, the collateral factor for rsETH in Aave was set relatively high, providing a certain buffer against price volatility—precisely the buffer the attacker exploited. Second, the attacker used multiple addresses to spread out the borrowing, causing the health of each individual position to appear normal while the overall risk exposure was enormous. In addition, Aave’s oracle was unable to capture the true execution price of rsETH in the DEXes within a short time; the time-weighted average price (TWAP) mechanism it relied on had a delay, resulting in liquidation triggers occurring later than the speed at which assets were withdrawn.

How does DeFi composability amplify the risk of a single protocol?

Composability is a core advantage of DeFi, but it also becomes an accelerant for risk propagation. In the KelpDAO attack, risk spread rapidly along the chain of “cross-chain bridge — re-staked token — lending protocol.” The vulnerability in the cross-chain bridge led to rsETH minting, and rsETH, as collateral, created borrowing capacity in Aave, ultimately converting the false value of assets into real liquidity extraction. This propagation mechanism has nonlinear characteristics: an attack cost of $5 million ultimately resulted in nearly $200 million in bad debt and more than $13 billion in TVL leaving. After the incident, market participants quickly withdrew liquidity from Aave and other lending protocols, further exacerbating capital flight. As of April 20, 2026, the TVL across the entire DeFi network fell from about $115 billion before the event to below $102 billion, with the evaporation amount exceeding $13 billion.

Who is fleeing behind the evaporation of $13 billion in TVL?

The rapid decline in TVL reflects market behavior across three layers. The first layer is the directly affected Aave protocol, where users proactively withdrew about $4.5 billion in liquidity to avoid assets being locked up or participating in liquidation. The second layer is the aggregators and leverage protocols that have capital interactions with Aave. Due to uncertainty in the underlying lending market, these protocols were forced to reduce positions or pause services, leading to a passive outflow of about $3.5 billion. The third layer is the spread of panic sentiment in the market: users withdrew assets from other lending and staking protocols that were not directly related, resulting in an overflow outflow of about $5 billion. Notably, the speed of this capital flight ranks among the top in DeFi history; within 72 hours, the TVL decline reached 11.3%. By asset category, outflows of ETH and stablecoins were the most significant, decreasing by about $4.8 billion and $5.2 billion, respectively.

Can DeFi insurance cover blind spots in attacks like this?

Existing DeFi insurance protocols have extremely limited coverage for this incident. Mainstream insurance protocols such as Umbrella typically limit coverage to direct fund losses caused by smart contract vulnerabilities, but the boundaries for claims regarding indirect bad debt caused by “inter-protocol risk transmission” are not clear. In the KelpDAO attack, Aave’s bad debt did not originate from a vulnerability in Aave’s own contracts, but from abnormal inputs from external protocols. Whether insurance should compensate for such “external input risks” currently lacks a unified standard in the industry. In addition, price depeg and liquidation failures caused by the attack are often categorized under exclusion clauses such as “market risk” or “operational risk.” As of April 20, 2026, multiple insurance protocols have stated that they are evaluating the scope of claims for this incident, but it is expected that most losses will not be covered by insurance. This blind spot highlights the limitations of DeFi insurance in the face of systemic risk scenarios.

Summary

The KelpDAO cross-chain bridge attack is one of the most severe DeFi security cases affecting the industry so far in 2026. With an attack cost of about $5 million, it triggered nearly $200 million in Aave bad debt and more than $13 billion in TVL evaporation. The core lessons from the incident include: the permission configuration of cross-chain bridges must be deeply bound to business logic; lending protocols need to strengthen the calibration of risk-control parameters for non-mainstream collateral assets; and the DeFi insurance system urgently needs to expand to cover systemic risk transmission scenarios. While composability improves capital efficiency, it also requires protocols to establish clearer risk-isolation mechanisms. For the industry, this incident is not the end; it is an important milestone in driving an upgrade of DeFi risk-control standards.

FAQ

Q: In the KelpDAO attack, who ultimately bears the $200 million bad debt in Aave?

A: The bad debt is first covered by the reserve funds of the Aave protocol. If reserves are insufficient, the protocol will gradually make up the shortfall through subsequent liquidation income, fee accumulation, and other sources. Some losses may ultimately be indirectly borne by Aave’s liquidity providers, depending on the community governance decision plan.

Q: Will this attack affect other cross-chain bridges that use LayerZero?

A: The LayerZero protocol itself is not vulnerable; the issue lies in KelpDAO’s configuration mistake in message verification. However, if other cross-chain bridges adopt similar non-strict permission verification mechanisms, they also face the risk of being attacked. It is recommended that relevant projects immediately audit the message verification logic for cross-chain communications.

Q: How should investors avoid similar DeFi composability risks?

A: Investors should pay attention to dependencies between protocols and avoid putting a large amount of assets into highly nested DeFi strategies. At the same time, they should prioritize protocols that have undergone multiple rounds of audits, have risk-isolation mechanisms in place, and have mature liquidation contingency plans. Diversifying asset storage across protocols built on different underlying architectures is also an effective risk-management approach.